EP2278283A1 - Method for correcting temperatures in a power measuring device and power measuring device - Google Patents

Abstract

The method involves transmitting a temperature-corrected output signal to a display unit and/or a processing unit. A corrective parameter is calculated based on a force measurement signal (mL) and/or a temperature measurement signal (mT) using a thermodynamic model. The corrective parameter is utilized to correct the output signal, where a temperature difference that exists between a system temperature and an ambient temperature (Te), or between system temperatures is characterized by the corrective parameter. An independent claim is also included for a force measuring apparatus comprising a temperature sensor.

Description

The present invention relates to a method for temperature correction of a force measuring device, in particular a balance, and a corresponding force measuring device.

In general, a force measuring device comprises various components such as a power take-up, a power transmission, a load cell and optionally a device for processing measurement signals. A force to be measured is absorbed by means of the force absorption and forwarded via the power transmission to the load cell. The force measuring cell converts the supplied force into an electrical force measuring signal which corresponds to the force acting on the force measuring device.

In a corresponding manner, the force acting on a balance is determined by the weight of the weighing sample, the so-called load, which acts on the force absorption in the form of a weighing pan. This force effect is transmitted via the power transmission in the form of a linkage to the load cell or load cell and there converted into an electrical force measurement signal, the so-called weighing signal.

The electrical force measurement signal is forwarded to a signal processing unit, which serves for the further processing of this force measurement signal and the generation of a corresponding output signal. The output signal is transmitted to a display unit and / or to a further processing device, for example to a host computer or to a system controller.

These force measuring devices or scales are typically used in weighing individual weighing samples but also in automated production and testing systems for weighing large quantities of weighing samples. The requirements of such a force measuring device consist in a high measurement accuracy, a high reproducibility and stability of the measurements. In addition, the force measuring device should be as simple and inexpensive as possible.

It is known that for accurate and stable measurements, interfering influences which distort the measurement result have to be measured and corrected in a suitable manner. For example, describes GB 1495278 a method in which the influences of load-independent parameters, in particular the influence of a temperature is corrected, which acts from the outside on the weighing device. For this purpose, the ambient temperature is measured by means of a temperature sensor and generates a corresponding electrical temperature measurement signal. On the basis of this temperature measuring signal, the force measuring signal is then processed to a temperature-corrected output signal. Time-dependent phenomena can also be corrected with this method, for example creeping by means of a time-dependent exponential function.

This known correction method is based on a state in which the components of the force measuring device have the same temperatures, namely the ambient temperature. Often, however, situations occur in which different components of the force measuring device have different temperatures. For example, with a force measuring device during operation, additional heat may be generated which results in an elevated temperature in some components of the force measuring device. Between these operating temperatures and the acting ambient temperature thus results in a temperature difference.

In general, such temperature differences affect the measurements of the force measuring device, ie the measurements of the applied force or load. Therefore, the aim is to correct the influence of these temperature differences as completely as possible, especially with high demands on the accuracy and stability of the force measuring device.

Various measures are known from the prior art to correct the effects of temperature differences. For example disclosed CH 658516 a balance based on the principle of electromagnetic force compensation. In this case, the compensation force is generated with the release of heat by means of a current-carrying coil, which is arranged in an air gap of a permanent magnet. Thus, a higher temperature results in the coil than in those components which are located further away from the coil are and whose temperature is determined essentially by the externally acting temperature. To compensate for this temperature difference, a temperature sensor is arranged in the interior of the permanent magnet and thus a temperature is measured which corresponds to the temperature of the heated permanent magnet. Based on this measured temperature then the force measurement signal is corrected. However, due to the thermal inertia of the permanent magnet, the temperature sensor can only react slowly to changes in the ambient temperature. Thus, despite the compensation errors occur which can falsify the measurement result. In addition, the installation, adjustment and control of the temperature sensor in the permanent magnet is complex and error-prone. These disadvantages result from the arrangement of the temperature sensor at or in a critical component, which is directly related to the actual force measurement.

For direct detection of the heat emission is in CH 669041 proposed to arrange the temperature sensor within the winding of the coil. Thereby, the temperature in the center of the heat source, namely within the winding of the coil, can be detected. However, even in this case, the installation and adjustment of the temperature sensor is problematic. Further, the measurement may not be representative of the correction of the temperature dependence of the magnetic field, as it is primarily related to the temperature of the permanent magnet and secondarily to the coil temperature.

The present invention is therefore an object of the invention to provide a method for temperature correction in a force measuring device, in particular a balance, with which a simple and cost-effective operation can be achieved while meeting high requirements in terms of measurement accuracy and stability. In particular, a method for compensating a temperature difference is to be specified, in which the temperature sensor can be easily installed, adjusted and controlled. In addition, a corresponding force measuring device should be given a simple, inexpensive and reliable construction.

This object is achieved by a method and with a force measuring device having the features specified in the independent claims. Advantageous embodiments of the invention are specified in the further, dependent claims.

The invention relates to a method for temperature correction of a force measuring device, in particular a balance, comprising the steps of: generating an electrical, corresponding to the applied force force measuring signal by means of a load cell; Measuring a temperature which mainly corresponds to an ambient temperature acting on the force measuring device and generating an electrical temperature measuring signal corresponding to the measured temperature; Processing the force measurement signal based on the temperature measurement signal and the force measurement signal to a temperature-corrected output signal; Transmission of the output signal to a display unit and / or to a further processing unit. In this case, in the step of processing from the force measurement signal and / or the temperature measurement signal, at least one correction parameter of the output signal is calculated, which characterizes a temperature difference between a system temperature and the measured temperature and / or between a first system temperature and a second system temperature consists. As a result, a simple, efficient and precise temperature compensation can be carried out without the need for a complicated and / or error-prone installation of a temperature sensor in or in critical components of the force measuring device. Thus, a simple and inexpensive construction of the force measuring device can be achieved. Furthermore, there are advantages in terms of cost, reliability and stability of the force measuring device, since the calculations can be carried out in simple steps in reasonable steps.

According to the invention, an additional thermal influence is thus calculated compared to the prior art, which significantly influences the force measurements. This calculated influence is then used to additionally correct the force measurement signal.

The ambient temperature is to be understood as meaning a temperature which comprises various ambient temperature influences which act on the force-measuring device, for example the temperature of the atmosphere surrounding the force-measuring device or which is located in the interior of a housing of the force-measuring device, as well as the temperature of the installation surface or of the table or the temperature of the material to be measured.

The ambient temperature acts on the force measuring device and on a temperature sensor arranged on or in the force measuring device. Thus, a temperature is determined with this temperature sensor, which corresponds in some way to the ambient temperature. Often, however, the measured temperature deviates from the ambient temperature, for example, with other temperature influences, which are related to the operation of the force measuring device.

System temperature is the temperature which can be assigned to a system component of the force measuring device, for example the temperature of a component or a part of a component or a group of components such as coil, permanent magnet, core of the permanent magnet, air gap, lever, position sensor or its coupling and Power input, power supply or load cell.

The temperature sensor is now arranged such that the measured temperature and thus the generated electrical temperature measurement signal mainly corresponds to the ambient temperature. For example, the sensor may be in direct thermal contact with either the surrounding atmosphere or those components whose system temperature is primarily determined by the ambient temperature. Less suitable is an arrangement of the temperature sensor at or in the vicinity of components whose temperature is determined essentially by operational processes of the force measuring device. Preferably, therefore, the temperature sensor is arranged at or in the vicinity of thermally passive or stationary components of the load cell. But it can also be exposed to the environment or external components of the force measuring device as on the housing or on moving parts such as the power or the power supply be attached. Furthermore, the temperature sensor could also be arranged outside the housing or in the vicinity of the force measuring device.

This arrangement has the advantage that the temperature sensor can react immediately and quickly to changes in the ambient temperature. In addition, the measurement of the ambient temperature is largely independent of the force measurement and accordingly of their influences, such as heat generation, decoupled. It is particularly advantageous that the temperature sensor can be arranged at almost any, non-critical locations and components of the force measuring device, so that the installation, adjustment and / or control of the temperature sensor can be easily made.

1. 1.) Bei einer Anordnung des Temperatursensors in thermischem Kontakt zu einem Bauteile kann die Temperaturdifferenz eine Differenz zwischen einer ersten gemessenen Systemtemperatur und einer zweiten berechneten Systemtemperatur sein.
2. 2.) Bei einer thermisch isolierten Anordnung des Temperatursensors kann sich die Temperaturdifferenz aus der Differenz zwischen der gemessenen einwirkenden Temperatur und einer berechneten Systemtemperatur ergeben.
3. 3.) Falls die gemessene Temperatur auch einer gemessen Systemtemperatur entspricht, kann die Temperaturdifferenz eine Kombination des Falls 1.) und des Falls 2.) sein.
4. 4.) Bei einer Differenzbildung zweier verschiedener Temperaturdifferenzen kann die zu charakterisierende Temperaturdifferenz als Differenz zwischen einer ersten berechneten Systemtemperatur und einer zweiten berechneten Systemtemperatur definiert sein.

In principle, the method according to the invention can be used for arbitrary corrections of temperature differences in force measuring devices. Thus, a temperature difference can be corrected, which results from a generated heat. However, it is also possible to correct temperature differences, which generally result from the difference between the ambient temperature and a system temperature or between two system temperatures. The correction parameter CP can thus characterize different temperature differences, for example:

1. 1.) In an arrangement of the temperature sensor in thermal contact with a component, the temperature difference may be a difference between a first measured system temperature and a second calculated system temperature.
2. 2.) In a thermally insulated arrangement of the temperature sensor, the temperature difference may be the difference between the measured applied temperature and a calculated system temperature.
3. 3.) If the measured temperature also corresponds to a measured system temperature, the temperature difference may be a combination of case 1.) and case 2.).
4. 4.) In a difference formation of two different temperature differences, the temperature difference to be characterized may be defined as the difference between a first calculated system temperature and a second calculated system temperature.

Situations with different system temperatures can occur when the ambient temperature changes relatively quickly and the system temperatures can not or at least not directly follow these changes due to the thermal inertia of the components. The resulting temperature difference is thus based on a short-term or abrupt change and therefore occurs only during a certain period of time, while the system temperature of a component with a time lag behind the jump of the ambient temperature. As a rule, however, this temperature difference largely equalises with time until, according to a certain reaction rate, a new equilibrium state is reached.

The reaction rate and thus the formation of a time-dependent temperature difference depends on the thermal characteristics of the components themselves and / or their connection to other components and / or to the environment. In this case, this characteristic can be associated with a variety of thermodynamic factors, for example, with the thermal inertia, heat, heat conduction and heat dissipation and with the mass or the surface of the components as well as the thermal temperature expansion.

Particularly advantageous in the inventive correction is that in principle there is no local restriction, since the system temperature can be calculated at any location and / or components of the force measuring device. Thus, there are no locally limited measuring ranges as with temperature sensors whose measured value may not be representative of the component to be measured.

Furthermore, the system temperature can be determined without additional expenditure for different locations and / or components without having to install a large number of sensors. Thus, even very complex forms of temperature distributions can be determined in a cost-effective manner and used for correction. In particular, based on a single unproblematic temperature measurement, the temperature difference between two different components of the power transmission, or their influence on the measurements, can be calculated.

Furthermore, thermal incorrect influences which arise due to a thermal connection of the temperature sensor are eliminated. For example, the adhesion of a temperature sensor to a component usually results in a heat-insulating intermediate layer which often leads to falsifications and / or delays in the system temperature to be measured. Such problems are not present when calculating the system temperature.

Furthermore, complex and varied temperature influences can be calculated efficiently and used to correct the force measurement signal by refining the calculations according to the accuracy requirements.

The correction parameter according to the invention should in principle be regarded as a theoretical calculation variable which is suitable for calculating the corresponding corrections. A correspondence with a physically existing physical quantity, for example a measurable temperature, is not absolutely necessary. However, the correction parameter should at least partially represent the characteristic temperature differences.

The term signal processing unit is understood to mean all signal processing elements which are suitable for processing the measuring signals of the load cell and of the temperature sensor, for example analog or digital circuits, integrated circuits, shift registers, processors, computers and the like.

In a first advantageous embodiment of the invention, the correction parameter is calculated by means of a response function and a convolution integral. As a result, thermodynamic phenomena which lead to corresponding temperature differences can be calculated in a compact form and used for correction. Preferably, the correction parameter is calculated by the temperature measurement signal, the force measurement signal or one of the force measurement signal determined power loss is folded with the time derivative of the response function. As a result, the measured signal, in particular its time lag, can be taken into account in the calculation of the correction parameter. The individual steps of this calculation are given in the detailed description.

In principle, a multiplicity of time-dependent functions can be used as the response function, for example exponential functions or polynomials. The response function is preferably defined according to at least one thermodynamic regularity and / or as a thermodynamic reaction of an incoming step function or impulse function and / or as a time-decreasing function, in particular as an exponential function with a predetermined time constant. As a result, various thermodynamic processes such as heat conduction thermal expansion or the thermal attenuation of a magnetic field can be mapped and used to calculate the correction parameter. Furthermore, the calculation can easily be adapted to the specific circumstances, for example by adapting the time constants to the thermal conductivity of the various materials.

In a preferred embodiment of the invention, the correction parameter is calculated by a recursive calculation method. This allows the computation of complex formulas, in particular the computation of the convolution integral, by means of simple mathematical operations, for example by means of simple multiplications, summations and delays. In addition, this method allows a strong reduction in the storage of numerical values and intermediate results. Furthermore, a largely delay-free signal processing or real-time processing is achieved.

Preferably, the correction parameter is calculated as a time-dependent variable, which depends on the time course, in particular before the time change, the measured temperature and / or the force measurement signal. By correcting these phenomena, temperature differences are primarily recorded that appear in the short term and / or temporarily. Thus, a so-called dynamic temperature correction can be made and thus a high measurement accuracy can be achieved even with short-term temperature changes. Particularly advantageous is the possibility of immediate consideration of the time dependence by the calculation. This delay-free correction of temperature differences is advantageous over temperature measurements, in which distortions by the temperature sensor itself can be expected, due to its thermal characteristics, in particular its thermal inertia.

In a preferred embodiment of the invention, the measured temperature is determined by the system temperature of a component of the force measuring device, in particular a component of the load cell, said system temperature mainly corresponds to the acting ambient temperature. As a result, the effect of the ambient temperature is measured directly at the corresponding components and thus made a direct reference to its temperature behavior. This results in an accurate detection of the effectively acting, component-specific temperature influences.

In a further embodiment, the system temperature corresponds to a temperature of a component of the force measuring device, in particular a force supply, a power transmission, a lever or a lever arm and the output signal is corrected in dependence on the temperature expansion of this component. As a result, a temperature difference, which leads to changes in the component dimensions to a fault influence on the force measurement, can be effectively corrected.

Basically, a temperature change causes a change in the component dimensions of all affected components. However, the dimensions of the affected components usually do not change at the same speed, but their temperature expansion depends on the individual characteristics of the components. In the case of these components and / or parts thereof, this leads to a change in the transmission of force or to a shift in the equilibrium. This results in an error influence on the force measurement, which, however, can be effectively corrected by the method according to the invention.

Preferably, the component is a lever or a lever arm, which serves to transmit the force of the applied force to the force cell. Thus, the correction parameter corresponds to a length extension of the lever or the Lever arms. As a result, the ambient temperature effects, as set forth in the detailed description, can be calculated using a simple thermodynamic model.

In a further exemplary embodiment according to the invention, the correction parameter characterizes at least two of the temperature differences which are respectively assigned to different components of the force-measuring device, in particular two counteracting lever arms. Such temperature differences occur when different components with different thermal characteristics respond to a short-term or abrupt temperature change in the ambient temperature. In this case, the inventive method is particularly advantageous because it can be dispensed with an elaborate processing of several temperature measuring signals and a correspondingly difficult adjustments. In addition, the installation of several temperature sensors is eliminated, resulting in a cost-effective design.

The method according to the invention is particularly advantageous in the case of complex structures with a variety of components, since a multiplicity of temperature differences can be characterized with the correction parameter according to the invention. Thus, complex thermal relationships can be detected and used for correction.

In particular, the correction parameter may comprise a group of two values comprising a first temperature difference of a first lever arm and a second temperature difference of a second lever arm. As a result, the complex behavior of several superimposed temporal processes, namely the different time courses of the two lever arms, can be calculated.

In a further embodiment according to the invention, the correction parameter is calculated by means of a difference signal, which results from the difference between two temperature differences. In this subtraction eliminates the dependence on a further reference temperature, in particular the acting ambient temperature and the temperature differences can be calculated as a single signal, namely as a difference signal. This simplifies the Handle the correction parameter, in particular its transfer to the further calculations.

In a further exemplary embodiment according to the invention, the force-measuring device is based on the principle of electromagnetic force compensation, and the correction parameter characterizes a temperature difference which is related to the generation of the compensation force. In this principle, a so-called compensation force is generated in the load cell, which counteracts the force acting on the force measuring device. Thus, the load cell always remains in the same position even with a change in weight, so that the phenomenon of creep initially described with reference to the prior art, so problems of deformation and / or material fatigue, can not or only to a very small extent occur.

The heat released in the generation of the compensation force generally leads to elevated temperatures of those components which serve to generate the compensation force. Thanks to the temperature correction according to the invention, this influence can now be corrected particularly simply by calculating the temperatures or their effects at the critical locations of the heating. Therefore, the time-consuming installation, adjustment and monitoring of temperature sensors on these components can be dispensed with. This is particularly advantageous because these components are usually difficult to access.

In a further embodiment, the compensation force is generated by means of a current-carrying coil to release a power loss, and calculates the correction parameter in dependence on the force measurement signal, in particular a power loss derived therefrom. This power loss causes a temperature change in the coil and the components surrounding the coil. This temperature change in turn causes a temporally variable force effect in the form of a so-called load drift. This load drift can now be corrected simply and precisely with the method according to the invention by setting the correction parameter directly as a function of the heat source, namely the power loss released in the coil.

Particularly advantageous is the inventive method in time-dependent processes in which the coil is traversed by a time-varying current whose current depends on a momentarily acting force. Thus, different power dissipation in the form of heat are released at different currents, resulting in different temperatures and thus temporally changed force effects. Through the calculation, these variable states can be compensated immediately and in a flexible manner.

In a further preferred embodiment of the invention, the compensation force is generated by the interaction of a coil and a permanent magnet and the system temperature corresponds to a temperature of the coil and / or the permanent magnet. Since both the coil and the permanent magnet are used to generate force, these components are particularly critical with regard to interfering interventions for measuring the system temperatures.

Since the coil is usually arranged in an air gap of the permanent magnet, a change in temperature of the coil by heat transfer via various components of the force measuring device or the air in the air gap results in a corresponding change in temperature of the permanent magnet. A change in temperature of the permanent magnet in turn influences its magnetic field and thus the compensation force generated. Thus, a constant acting load leads to an increasing warming, thus to an apparent time-dependent change of the acting load, thus to the phenomenon of the load drift.

In the inventive calculation of the system temperature of the permanent magnet, this heat transfer between the causative generation of temperature changes in the coil and their effect on the magnet can now be calculated in a good approximation and thus used for temperature correction.

In a further preferred embodiment, at least two corrections of the force measuring signal are calculated by at least two different, substantially separate modules, wherein in each case at least one of the mentioned correction parameters is calculated by each module. Due to this modular design, the Calculation can be put together on the most diverse temperature influences and adapted in a simple way to changing needs.

Furthermore, the invention relates to a force measuring device, in particular a balance, with a load cell, which generates an electrical, corresponding to the applied force force signal, with a temperature sensor, which measures a temperature that mainly corresponds to an acting on the force measuring device ambient temperature and a measured Temperature corresponding electrical temperature measurement signal generated, and with a signal processing unit which is configured to process the force measurement signal based on the temperature measurement signal and the force measurement signal to a temperature-corrected, transferable to a display unit and / or to another processing unit output signal. In this case, the signal processing unit is configured such that a correction parameter serving to correct the output signal can be calculated, which characterizes a temperature difference between a system temperature and the measured temperature and / or
a first system temperature and a second system temperature.

In a further embodiment of the force measuring device, the temperature sensor is in thermal contact with a, in particular stationary, component of the force measuring device, wherein the system temperature of the component is determined primarily by the acting ambient temperature. As a result, the influence of the ambient temperature on the components can be detected directly and thus a precise compensation of these influences can be achieved. Preferably, the temperature sensor is arranged on the stationary components of the load cell, for example, at the attachment, Thus, the temperature sensor can be easily mounted on the unproblematic components of the force measuring device and the temperature measurement signal are passed without affecting the force measurement.

Preferably, the method according to the invention is carried out with a program which can be executed in a signal processing unit and which serves to calculate the output signal according to the method according to the invention. As a result, a high degree of flexibility and reusability of the calculation method can be achieved.

Fig. 1

eine Waage 1 in beispielhafter Ausgestaltung mit symbolisch dargestellter Last L sowie mit einer von aussen einwirkenden Umgebungstemperatur Te und einer innerhalb der Kraftmessvorrichtung wirkenden Systemtemperatur Ts;

Fig. 2

eine vereinfachte schematische Zeichnung einer Kraftmesszelle 10, welche auf dem Prinzip der elektromagnetischen Kraftkompensation basiert, für eine Waage gemäss Figur 1 im Schnitt, welche ein Kraftmesssignals mL erzeugt und einen Temperatursensor 15 aufweist, der eine Temperatur Tm misst, die hauptsächlich zur Umgebungstemperatur Te korrespondiert und ein dazu entsprechendes Temperaturmesssignal mT erzeugt;

Fig. 3

das Blockschaltbild der Waage 1 gemäss Figur 1 und 2, mit einer Signalverarbeitungseinheit 20, welche anhand des Kraftmesssignals mL, des Temperaturmesssignals mT und eines Korrekturparameters CP ein temperaturkorrigiertes Ausgabesignal mLc erzeugt, das an eine Anzeigeeinheit 13 oder eine weitere Verarbeitungseinheit 14 weitergeleitet wird;

Fig. 4

ein Blockschaltbild gemäss Figur 3, mit einer ersten Temperaturkorrektur durch einen ersten Korrektursummanden C_S und einer erfindungsgemässen Korrektur durch einen zweiten Korrektursummanden C_LD, wobei der Korrekturparameter CP als relative Temperaturdifferenz dTr ausgebildet ist, und der zweite Korrektursummanden C_LD in Abhängigkeit von dieser relativen Temperaturdifferenz dTr berechnet wird;

Fig. 5a

ein weiteres Ausführungsbeispiel der erfindungsgemässen Berechnung zweier Temperaturdifferenzen dT1 und dT2, mit einem Blockschaltbild entsprechend Figur 4;

Fig. 5b

einem Schema zur Herleitung der Berechnungsmethode für das Ausführungsbeispiel gemäss Figur 5a;

Fig. 5c

schematisch dargestellte Temperaturverläufe bei einem Temperatursprung dTe der Umgebungstemperatur Te für das Ausführungsbeispiel gemäss Figur 5a;

Fig. 6

ein Blockschaltbild, bei welchem die Berechnungen gemäss Figuren 4, 5a, 5b, und 5c als separat arbeitende Berechnungsmodule kombiniert sind.

Details of the inventive method and the corresponding force measuring device will become apparent from the description of the embodiments illustrated in the drawings. Show it:

Fig. 1

a balance 1 in an exemplary embodiment with a symbolically represented load L and with an ambient temperature Te acting from outside and a system temperature Ts acting within the force measuring device;

Fig. 2

a simplified schematic drawing of a load cell 10, which is based on the principle of electromagnetic force compensation, for a balance according to FIG. 1 in section, which generates a force measuring signal mL and has a temperature sensor 15, which measures a temperature Tm which corresponds mainly to the ambient temperature Te and generates a corresponding temperature measurement signal mT;

Fig. 3

the block diagram of the balance 1 according FIGS. 1 and 2 with a signal processing unit 20, which generates a temperature-corrected output signal mLc based on the force measurement signal mL, the temperature measurement signal mT and a correction parameter CP, which is forwarded to a display unit 13 or a further processing unit 14;

Fig. 4

a block diagram according to FIG. 3 with a first temperature correction by a first correction sum C_S and a correction according to the invention by a second correction sum C_LD, wherein the correction parameter CP is designed as a relative temperature difference dTr, and the second correction sum C_LD is calculated as a function of this relative temperature difference dTr;

Fig. 5a

a further embodiment of the inventive calculation of two temperature differences dT1 and dT2, with a block diagram corresponding FIG. 4 ;

Fig. 5b

a scheme for deriving the calculation method for the embodiment according to FIG. 5a ;

Fig. 5c

shown schematically temperature profiles at a temperature jump dTe the ambient temperature Te for the embodiment according to FIG. 5a ;

Fig. 6

a block diagram in which the calculations according to FIGS. 4 . 5a, 5b, and 5c are combined as separate calculation modules.

FIG. 1 shows a scale 1 in an exemplary embodiment with a symbolically represented load L, which acts on the balance. In this case, a compensation force F is generated, which counteracts the acting load L. Furthermore, an ambient temperature Te is shown symbolically, which acts on the balance from the outside and which significantly influences the accuracy and stability of the measurements. On the one hand, the measured values ascertained by the balance change as a function of the load L, but on the other hand these measured values also change as a function of the ambient temperature Te and through the action of a system temperature Ts acting within the force measuring device.

As described in the following sections, this system temperature Ts can deviate from the ambient temperature Te by various causes. For example, when generating the compensation force, a loss of heat in the interior of the balance and thus an increased system temperature Ts. Temperature differences can also arise due to the thermal inertia of the components of the force measuring device, if the temperature changes of the components can not follow the applied temperature changes sufficiently quickly. It is therefore the task of the signal processing unit provided in the scale 1 to compare the measured values corresponding to the load L as well as possible from these disturbing ones To separate temperature effects and thus to achieve accurate and stable measurements on the display 13, such as a liquid crystal display.

FIG. 2 shows very schematically a suitable for use in weighing technology load cell 10, which is based on the principle of electromagnetic force compensation, on average. The force measuring cell 10 has various system components or components, namely a device for power transmission with a parallel guide with a fixed portion 42 and a vertically deflectable portion 43, which are each hinged at the ends of a pair of links 44 via Biegelager 45. The vertically deflectable region 43 has a cantilever 41, which serves to receive an acting, to be measured load L, which is shown schematically by an arrow. The normal component of the load L, caused force is transmitted from the vertically deflectable portion 43 by a coupling element 49 on the first lever arm 48 of the lever 46. The lever 46 is supported by a flexure bearing 47 on a part of the fixed portion 42. The load cell 10 further comprises a cup-shaped permanent magnet 50, which is arranged in fixed connection with the fixed part 42 and has an air gap 51. In the air gap 51 a connected to the second lever arm of the lever 46 coil 53 is arranged. The coil 53 is traversed by a compensation current Ic, the size of which depends on the force acting on the lever 46. The position of the lever 46 is measured by an opto-electric position sensor 54 which is connected to a control device 56. The control device 56 controls the compensation current Ic as a function of the supplied position-measuring signals such that the lever 46 is always kept in the same position or is returned after a change in the load L in this again. The control device 56 generates a digital or analog force measuring signal mL, which is forwarded for further signal processing.

Depending on the different acting force, a correspondingly different compensation current Ic must be conducted through the coil 53. This results in a load-dependent power loss in the coil 53. This coil 53 thus acts as a load-dependent heat source and thus leads to a corresponding change in temperature of the affected components. This temperature change pertains especially to the coil 53 itself, but also to those heated by heat transfer Components, such as the permanent magnet 50, the lever 46 and the position sensor 54th

Furthermore, a temperature sensor 15 is arranged on the stationary part 42 of the load cell 10. As a result, with this temperature sensor 15 a temperature Tm is measured, which corresponds to a system temperature of the fixed part 42. Since the fixed part 42 is in direct thermal contact with the acting ambient temperature Te, a temperature Tm is measured which corresponds indirectly to the environmental temperature Te acting on the balance 1.

However, the temperature sensor 15 is disposed at some distance from the heat generating coil 53. Thus, the temperature of the coil 53 has only a minor influence on the measured temperature Tm. Consequently, the temperature sensor 15 mainly measures the effect of the ambient temperature Te on the balance 1.

However, the temperature sensor 15 can also be arranged remotely from the load cell 10 or other components or system components of the balance 1, so that the measured temperature Tm is largely independent of a system temperature Ti of the components of the balance 1. For example, the temperature sensor 15 could also be arranged on a housing wall of the balance 1.

The temperature sensor 15 generates a digital or analog temperature measurement signal mT, which corresponds to the system temperature Ts of the fixed part 42 and thus mainly to the ambient temperature Te. This temperature measurement signal mT is then also, as the force measurement signal mL, forwarded for further signal processing.

FIG. 3 shows an exemplary representation of the block diagram of in FIGS. 1 and 2 shown scale 1. The load cell 10 generates in this example, an analog force measurement signal mL ', which corresponds to the force acting on the load cell 10 load L. The load cell 10 is connected to an analog-to-digital converter 11a to pass the analog measurement signal thereto. The analog-to-digital converter 11a converts the supplied force measurement signal mL 'into a digital force measurement signal mL, wherein this digital force measurement signal mL also corresponds to the applied load L (analog signals are referred to in this figure as apostrophe, the corresponding digital signal, however, without apostrophe). The output of the analog-to-digital converter 11a is connected to a first input of a signal processing unit 20 in order to forward the force measuring signal mL thereto.

Furthermore, the balance comprises the above-described temperature sensor 15 for detecting the measured temperature Tm. This temperature sensor 15 is connected via a further analog-to-digital converter 11b to the second input of the signal processing unit 20 in order to provide the temperature measurement signal mT 'generated by the temperature sensor 15, after conversion into digital form, to the signal processing unit 20 as a further input variable.

In this example, separate analog-to-digital converters 11a and 11b are used for the force measurement signal mL and for the temperature measurement signal mT, since often the requirements with respect to resolution and speed of the measured signals differ greatly. However, the analog-to-digital converters 11a and 11b can also be combined in one unit. Furthermore, the analog-to-digital converters 11a and 11b can also be integrated in the force measuring cell 10, or in the temperature sensor 15 or in the signal processing unit 20.

In the arithmetic unit 20, the original force measurement signal mL is processed on the basis of the temperature measurement signal mT to form a temperature-corrected output signal mLc. The temperature-corrected output signal mLc corresponds approximately, that is, apart from the corrected temperature errors, to the original force measurement signal mL and thus also to the applied load L. The output of the signal processing unit 20 is connected to a display unit 13, so that the corrected output signal mLc forwarded to this and there can be displayed. The corrected output signal mLc can also be forwarded to a further processing unit 14, for example to a monitoring / alarming device or to a host computer or to a process controller.

$$\mathrm{mLc}\mathrm{=}\mathrm{F}\left(\mathrm{CP}\mathrm{mL}\mathrm{mT}\right)\mathrm{.}$$

In the signal processing unit 20, a correction parameter CP is calculated from the force measurement signal mL and from the temperature measurement signal mT by means of a mathematical function. This correction parameter CP characterizes, as described in the following sections, a temperature difference which exists between a system temperature and the measured temperature Tm and / or a first system temperature and a second system temperature and serves to generate the temperature-corrected output signal mLc. In a further step, the force measuring signal mL is processed by means of the correction parameter CP to form a temperature-corrected output signal mLc. These arithmetic operations can be represented in general form by the following formulas: $$\mathrm{CP}\mathrm{=}\mathrm{F}\left(\mathrm{mL}\mathrm{mT}\right)$$

$$\mathrm{mLc}\mathrm{=}\mathrm{F}\left(\mathrm{CP}\mathrm{mL}\mathrm{mT}\right)\mathrm{,}$$

In this embodiment, the correction parameter CP characterizes two different temperature differences: On the one hand, a temperature difference results from the difference between the system temperature Ts of the heated ones FIG. 3 described coil 53 and the measured temperature Tm. On the other hand, the temperature sensor 15 according to the above also measures the temperature of the fixed area 43. Thus, the correction parameter CP characterizes a further temperature difference, which is between a first system temperature Ts1, ie the temperature of the coil 53, and a second system temperature Ts2, namely the temperature of the fixed area 43 results.

In another embodiment, such as for example in the following FIG. 5 described, the correction parameter CP can also characterize temperature differences, as they occur between different system temperatures of two different components in the case of temperature jumps of the acting ambient temperature.

The above-described blocks, namely the analog-to-digital converters 11a and 11b and the digital processing unit 20 are preferably implemented in a functional unit and suitably combined (dashed line), for example as Assemblies on a common board, in a common housing or in a microprocessor.

FIG. 4 shows a block diagram according to FIG. 3 , but with a combination consisting of a first temperature correction by a first correction sum C_S and a correction according to the invention by a second correction sum C_LD, which by means of the FIG. 3 calculated correction parameter CP is calculated. In this case, the correction parameter CP characterizes a relative time-dependent temperature difference dTr (t), which results from the temperature difference between the system temperature Ts of the coil 53 and the measured temperature Tm.

The original signals, the measurement signal mL 'of the compensation current Ic and the temperature measurement signal mT' of the measured temperature Tm are supplied to the analog-to-digital converter 11 and converted there into digital form. The corresponding output signals, namely the force measurement signal mL and the temperature measurement signal mT are then available in digital form for the following calculations.

1. a) Das Magnetfeld im Permanentmagnet 50 hat eine relativ grosse Temperaturabhängigkeit mit einem Temperaturkoeffizient von -350 ppm/K für SmCo. Folglich ist auch die Stromstärke in der Spule 53 relativ stark von der Temperatur abhängig. Das Kraftmesssignal mL korrespondiert wiederum mit der Stromstärke in der Spule 53. Daher wird, wie aus dem Stand der Technik bekannt, in einer ersten Korrektur das Kraftmesssignal mL bezüglich der Temperaturabhängigkeit des Permanentmagneten 50 bezüglich der gemessenen Umgebungstemperatur Te kompensiert. Dementsprechend wird diese Korrektur auch als statische Temperaturkorrektur oder als Temperaturkorrektur erster Ordnung bezeichnet. Dabei wird aus dem Kraftmesssignal mL und aus dem Temperaturmesssignal mT der erste Korrektursummand C_S gemäss folgender Formel berechnet: $$\mathrm{C\_S}=\mathrm{mL}*\mathrm{TK}*\mathrm{mT}$$
   
   
   wobei die Grösse TK ein konstanter, typischerweise empirisch ermittelter Wert, nämlich der Temperaturkoeffizient TK ist. Der erste Korrektursummand C_S wird in diesem Beispiel in einfacher linearer Abhängigkeit aus dem Kraftmesssignal mL und dem Temperaturmesssignal mT berechnet. Der erste Korrektursummand C_S kann aber auch unter Berücksichtigung höherer Terme, zum Beispiel dem quadrierten Kraftmesssignal (mL)², berechnet werden.
2. b) Der berechnete erste Korrektursummand C_S wird einem Summenoperator zugeführt, der durch eine Addition des ersten Korrektursummand C_S zum ursprünglichen Kraftmesssignal mL eine entsprechende Korrektur durchführt. Das Kraftmesssignal mL wird somit durch eine erste Temperaturkorrektur gegenüber Änderungen der einwirkenden Last L und gegenüber Änderungen der gemessenen Temperatur Tm korrigiert. In gleicher Weise wie in Figur 3 wird das korrigierte Messsignal schliesslich an eine Anzeigeeinheit 13 und/oder an eine weitere Verarbeitungseinheit 14 weitergeleitet.

The following steps essentially correspond to the known steps according to the prior art, for example, those in GB 1495278 disclosed methods:

1. a) The magnetic field in the permanent magnet 50 has a relatively large temperature dependence with a temperature coefficient of -350 ppm / K for SmCo. Consequently, the current in the coil 53 is relatively strongly dependent on the temperature. The force measurement signal mL in turn corresponds to the current in the coil 53. Therefore, as is known from the prior art, in a first correction, the force measurement signal mL with respect to the temperature dependence of the permanent magnet 50 with respect to the measured ambient temperature Te compensated. Accordingly, this correction is also referred to as static temperature correction or as first order temperature correction. In this case, the first correction sum C_S is calculated from the force measurement signal mL and from the temperature measurement signal mT according to the following formula: $$\mathrm{C\_S}=\mathrm{mL}*\mathrm{TK}*\mathrm{mT}$$
   
   
   wherein the quantity TK is a constant, typically empirically determined value, namely the temperature coefficient TK. The first correction sum C_S is calculated in this example in a simple linear dependence from the force measurement signal mL and the temperature measurement signal mT. However, the first correction sum C_S can also be calculated taking into account higher terms, for example the squared force measurement signal (mL) ² .
2. b) The calculated first correction sum C_S is fed to a sum operator, which performs a corresponding correction by adding the first correction sum C_S to the original force measurement signal mL. The force measurement signal mL is thus corrected by a first temperature correction with respect to changes in the applied load L and with respect to changes in the measured temperature Tm. In the same way as in FIG. 3 the corrected measurement signal is finally forwarded to a display unit 13 and / or to a further processing unit 14.

According to a preferred embodiment of the invention is now additionally from the force measuring signal mL a, FIG. 3 corresponding correction parameter CP calculated in accordance with the detailed description of the next sections. From this correction parameter CP, an additional second correction sum C_LD is calculated in a downstream calculation. On the basis of this second correction sum C_LD, the force measurement signal mL is additionally corrected. Thus, the correction according to the invention makes a higher-order temperature correction.

The correction parameter CP as a relative time-dependent temperature difference dTr (t) characterizes in this embodiment, the same characteristic time course as a principle measurable temperature difference and differ from this only by appropriate normalization constants. In general, however, the temperature difference is only an arithmetic variable, which does not necessarily have to be a real existing temperature difference.

wobei c eine Proportionalitätskonstante ist. In anderen Fällen, bei denen keine Proportionalität vorliegt, kann die Verlustleistung P(t) ebenfalls mit einfachen Berechnungen aus dem Kraftmesssignal mL ermittelt werden.In a first calculation, a time-dependent power loss P (t) is determined from the force measurement signal mL, which represents the time profile of the power loss in the coil 53. In this example, the force measuring signal mL is proportional to the coil current Ic, so that the following applies to the power loss P (t): $$\mathrm{P}\left(\mathrm{t}\right)\mathrm{=}\mathrm{c}\mathrm{*}{\mathrm{mL}}^{\mathrm{2}}$$

where c is a proportionality constant. In other cases in which there is no proportionality, the power loss P (t) can also be determined with simple calculations from the force measurement signal mL.

Furthermore, a time-dependent response function S (t) is used to calculate the relative temperature difference dTr (t). The response function S (t) is based on a thermodynamic regularity which characterizes the time-dependent behavior of a temperature difference of the balance 1. Therefore, the response function S (t) results from a law of thermodynamics, namely from the heat flow and / or the heat conduction within the balance 1.

The time-dependent relative temperature difference dTr (t) is then calculated by a mathematical operation of convolution of the power loss P (t) with the derived response function S (t): $$\mathrm{dTr}\left(\mathrm{t}\right)\mathrm{=}\mathrm{P}\left(\mathrm{t}\right)\mathrm{\cdot }\mathrm{S}\left(\mathrm{0}\right)\mathrm{+}\underset{\mathrm{0}}{\overset{\mathrm{t}}{\mathrm{\int }}}\mathrm{P}\left(\mathrm{t\; \text{'}}\right)\mathrm{\cdot }\frac{\mathrm{.DELTA.S}\left(\mathrm{t}\mathrm{-}\mathrm{t\; \text{'}}\right)}{\mathrm{\delta }\left(\mathrm{t}\mathrm{-}\mathrm{t\; \text{'}}\right)}\mathrm{dt\text{'}}\mathrm{,}$$

mit einer charakteristischen Zeitkonstanten τ. In diesem Fall berechnet sich dann die relative Temperaturdifferenz dTr(t) gemäss der obigen Formel zu: $$\mathrm{dTr}\left(\mathrm{t}\right)\mathrm{=}\frac{\mathrm{1}}{\mathrm{\tau }}\underset{\mathrm{0}}{\overset{\mathrm{t}}{\mathrm{\int }}}\mathrm{P}\left(\mathrm{mL}\left(\mathrm{tʹ}\right)\right)\mathrm{\cdot }{\mathrm{e}}^{\mathrm{-}\frac{\mathrm{t}\mathrm{-}\mathrm{tʹ}}{\mathrm{\tau }}}\mathrm{dtʹ}\mathrm{.}$$

To determine the response function S (t), a simple thermodynamic model is given as an example. Accordingly, the response function S (t) results as a so-called step response from a stepped change in the force measurement signal mL. In this case, the step-shaped change forms an input variable on the balance 1, which generates a corresponding response function in accordance with a thermodynamic regularity, in this case according to the heat conduction in the components of the force-measuring device. According to this law, the time-dependent response function S (t) results according to the following formula: $$\mathrm{S}\left(\mathrm{t}\right)\mathrm{=}\mathrm{1}\mathrm{-}{\mathrm{e}}^{\mathrm{-}\mathrm{t}\mathrm{/}\mathrm{\tau }}$$

with a characteristic time constant τ. In this case, the relative temperature difference dTr (t) is then calculated according to the above formula: $$\mathrm{dTr}\left(\mathrm{t}\right)\mathrm{=}\frac{\mathrm{1}}{\mathrm{\tau }}\underset{\mathrm{0}}{\overset{\mathrm{t}}{\mathrm{\int }}}\mathrm{P}\left(\mathrm{mL}\left(\mathrm{t\; \text{'}}\right)\right)\mathrm{\cdot }{\mathrm{e}}^{\mathrm{-}\frac{\mathrm{t}\mathrm{-}\mathrm{t\; \text{'}}}{\mathrm{\tau }}}\mathrm{dt\text{'}}\mathrm{,}$$

For modeling, other response functions S (t) or step responses can also be used, in particular step responses which have a combination of several exponential functions.

wobei die Summe Σ_P(n) durch die Rekursionsformel $$\mathrm{\sum }_{\mathrm{P}}\left(\mathrm{n}\right)\mathrm{=}\mathrm{c}\mathrm{\sum }_{\mathrm{P}}\left(\mathrm{n}\mathrm{-}\mathrm{1}\right)+\mathrm{0.5}\left(\mathrm{c\; P}\left(\mathrm{n}\mathrm{-}\mathrm{1}\right)\mathrm{+}\mathrm{P}\left(\mathrm{n}\right)\right)$$

mit $$\mathrm{c}\mathrm{=}{\mathrm{e}}^{\mathrm{-}{\mathrm{t}}_{\mathrm{s}}\mathrm{/}\mathrm{\tau }}\mathrm{<}\mathrm{1}$$

gegeben ist.Various methods are available for calculating the convolution integral of the time-dependent relative temperature difference dTr (t). For example, at a constant sampling rate with the sampling time t _s, the convolution integral can be approximated by a sum: $$\mathrm{dTr}\left(\mathrm{t}\mathrm{=}{\mathrm{nt}}_{\mathrm{s}}\right)\mathrm{=}\left({\mathrm{t}}_{\mathrm{s}}\mathrm{/}\mathrm{\tau }\right)\underset{\mathrm{P}}{\mathrm{\Sigma }}\left(\mathrm{n}\right)$$

where the sum Σ _P (n) by the recursion formula $$\underset{\mathrm{P}}{\mathrm{\Sigma }}\left(\mathrm{n}\right)\mathrm{=}\mathrm{c}\underset{\mathrm{P}}{\mathrm{\Sigma }}\left(\mathrm{n}\mathrm{-}\mathrm{1}\right)+\mathrm{0.5}\left(\mathrm{c\; P}\left(\mathrm{n}\mathrm{-}\mathrm{1}\right)\mathrm{+}\mathrm{P}\left(\mathrm{n}\right)\right)$$

With $$\mathrm{c}\mathrm{=}{\mathrm{e}}^{\mathrm{-}{\mathrm{t}}_{\mathrm{s}}\mathrm{/}\mathrm{\tau }}\mathrm{<}\mathrm{1}$$

given is.

These arithmetic operations can be carried out in various ways, for example as a computer program or as a recursive circuit which has delay units and / or sum operators and / or multiplication operators. This allows the convolution integral to be calculated efficiently and with high accuracy using simple mathematical operations. Thus, at the end of this calculation block, the correction parameter CP is available as a relative temperature difference dTr (t). This temperature difference dTr (t) would be determined in the prior art by appropriate measurements with a temperature sensor.

wobei die Grösse TKLD ein vorgegebener, konstanter Temperaturkoeffizient ist. Bei der Berechnung handelt es sich um einen sehr einfachen Fall. Durch Berücksichtigung höherer Terme kann jedoch die Genauigkeit der Berechnung gemäss vorgegebenen Anforderungen verbessert werden.In a further calculation, from the relative temperature difference dTr (t) the second correction sum C_LD is calculated according to the following formula: $$\mathrm{C\_LD}\mathrm{=}\mathrm{mL}\mathrm{*}\mathrm{TKLD}\mathrm{*}\mathrm{dTr}\left(\mathrm{t}\right)$$

where the quantity TKLD is a predetermined, constant temperature coefficient. The calculation is a very simple case. By taking into account higher terms, however, the accuracy of the calculation can be improved according to predetermined requirements.

The second correction sum C_LD is supplied, as in the case of the first correction, to the sum operator, which makes a corresponding second correction by additional addition of the second correction sum C_LD. The force measurement signal mL is thus corrected with respect to the relative temperature difference dTr (t) and the consequent load drift, which results from the generation of the compensation force. The lastdriftkorrigierte output signal mLc is then available at the output of the signal processing unit 20.

Figures 5a, 5b and 5c show a further embodiment of the inventive calculation of the correction parameter CP, which the time-dependent, transient temperature behavior of the balance according to Figures 1 and 2 characterized, namely the dynamic temperature behavior of the balance 1 with two lever arms 48a, 48b with temporal temperature changes of the ambient temperature Te in the form of a so-called temperature jump dTe.
In this case, the correction parameter CP characterizes a group of two temperature differences dT1 and dT2, which respectively result from the difference between the system temperature Ts1, Ts2 of the respective lever arm 48a, 48b and the measured temperature Tm. However, as described in the following sections, the correction parameter CP can also characterize a difference signal ΔdT, which in turn results from the difference between the two temperature differences dT1 and dT2.

The ambient temperature Te acts on a lever 46 in the force measuring device 1, and thus on the two lever arms 48a and 48b. Consequently, a first temperature difference dT1 results between the first system temperature Ts1 of the first lever arm 48a and the measured temperature Tm, and a second temperature difference dT2 between the second system temperature Ts2 of the second lever arm 48b and the measured temperature Tm.

FIG. 5a shows a block diagram according to FIG. 4 , but with an inventive calculation of two temperature differences dT1 and dT2. Thus, the correction parameter CP comprises two temperature differences dT1 and dT2, each of which is derived from the in FIGS. 3 and 4 described temperature measurement signal mT be calculated. In this case too, the calculated temperature differences dT1 and dT2 are not necessarily actually existing quantities in this example, but are to be regarded as arithmetic variables, as in the preceding exemplary embodiment. The temperature differences dT1 and dT2 are calculated from the following derivation.

FIG. 5b shows a highly simplified schematic diagram of a balance according to FIG. 4 with a lever 46 which serves to transmit the applied load L to the coil 53. This lever 46 is mounted in the pivot point X, the so-called fulcrum, movable. The lever 46 has a first lever arm 48a and a second lever arm 48b, which are respectively represented by the respective masses m1 and m2.

The center of gravity of the first lever arm 48a with the mass m ₁ has the distance a ₁ from the pivot point X of the lever device. This mass corresponds to the mass of the vertically deflectable portion 43 of the balance. The second lever arm 48b with the mass m2 corresponds to the mechanism for transmitting power in the balance 1 and has the center of gravity distance a ₂ . Thus, without an applied load L, the equilibrium condition results: $${\mathrm{m}}_{\mathrm{1}}{\mathrm{a}}_{\mathrm{1}}\mathrm{=}{\mathrm{m}}_{\mathrm{2}}{\mathrm{a}}_{\mathrm{2}}$$

$${\mathrm{a}}_{\mathrm{2}}\mathrm{=}\mathrm{\beta }{\mathrm{b}}_{\mathrm{2}}$$

wobei die Hebeluntersetzung gleich b₂/a₁ ist.In the case of a load, the load L, which is compensated at the end of the right-hand side by the magnetic force f _M, additionally acts on the left-hand side. It follows from the assumption that the ratio of the center of gravity distance a ₂ to the length b _{2 of} the right-hand side is constantly equal to β: $$\mathrm{L}{\mathrm{a}}_{\mathrm{1}}\mathrm{=}{\mathrm{f}}_{\mathrm{M}}{\mathrm{b}}_{\mathrm{2}}$$

$${\mathrm{a}}_{\mathrm{2}}\mathrm{=}\mathrm{\beta }{\mathrm{b}}_{\mathrm{2}}$$

where the lever ratio is equal to b ₂ / a ₁ .

With a temperature jump dTe, the two masses m ₁ and m _{2 are} heated. The mass m ₁ is relatively heavy and compact and thus the ratio of the surface to the mass is small. Therefore, the first system temperature Ts1 of the mass m1 will adapt only slowly to the temperature jump dTe. In contrast, the mass m _{2 is} light and has a relatively large surface area. Their temperature, ie the second system temperature Ts2, will thus follow the temperature jump dTe relatively quickly. This results for the lever arms 48a and 48b for a certain time different system temperatures Ts1 and Ts2 and thus also different temperature differences with respect to the ambient temperature Te, namely dT1 for the first lever arm 48a and dT2 for the second lever arm 48b.

Consequently, the expansion coefficient α results in different changes in length on both sides of the lever 46, namely
α * dT1 for the first lever arm 48a and
α * dT2 for the second lever arm 48b.

$${\mathrm{m}}_{\mathrm{1}}{\mathrm{a}}_{\mathrm{1}}\mathrm{=}{\mathrm{m}}_{\mathrm{2}}{\mathrm{b}}_{\mathrm{2}}$$

$${\mathrm{a}}_{\mathrm{2}}\mathrm{=}\mathrm{\beta }{\mathrm{b}}_{\mathrm{2}}$$

Consequently, after the temperature jump dTe, it shifts the mechanical balance according to the formula: $$\left({\mathrm{<figure-callout\; id="1"\; label="m"\; filenames="imgf0001.png"\; state="\{\{state\}\}">m</figure-callout>}}_{\mathrm{1}}\mathrm{+}\mathrm{L}\right){\mathrm{a}}_{\mathrm{1}}\left(\mathrm{1}\mathrm{+}\mathrm{\alpha \; dT}\mathrm{1}\right)\mathrm{=}{\mathrm{m}}_{\mathrm{2}}{\mathrm{a}}_{\mathrm{2}}\left(\mathrm{1}\mathrm{+}\mathrm{\alpha \; dT}\mathrm{2}\right)\mathrm{+}{\mathrm{f}}_{\mathrm{M}}{\mathrm{b}}_{\mathrm{2}}\left(\mathrm{1}\mathrm{+}\mathrm{\alpha \; dT}\mathrm{2}\right)$$

$${\mathrm{m}}_{\mathrm{1}}{\mathrm{a}}_{\mathrm{1}}\mathrm{=}{\mathrm{m}}_{\mathrm{2}}{\mathrm{b}}_{\mathrm{2}}$$

$${\mathrm{a}}_{\mathrm{2}}\mathrm{=}\mathrm{\beta }{\mathrm{b}}_{\mathrm{2}}$$

As in the previous embodiment according to FIG. 4 the temperature differences dT1 and dT2 are each calculated with a convolution integral and the time-dependent response functions S1 (t) and S2 (t) and the characteristic time constants τ ₁ and τ ₂ .

mit i=1 und 2. Die Berechnung der Temperaturendifferenzen dT1 und dT2 unterscheidet sich also nur in den verschiedenen Zeitkonstanten τ₁ und τ₂.Accordingly, for the respective folding integrals: $${\mathrm{dT}}_{\mathrm{i}}\left(\mathrm{t}\right)\mathrm{=}{\mathrm{mT}}_{\mathrm{i}}\left(\mathrm{t}\right)\mathrm{\cdot }{\mathrm{S}}_{\mathrm{i}}\left(\mathrm{0}\right)\mathrm{+}\underset{\mathrm{0}}{\overset{\mathrm{t}}{\mathrm{\int }}}{\mathrm{mT}}_{\mathrm{i}}\left(\mathrm{t\; \text{'}}\right)\frac{{\mathrm{.DELTA.S}}_{\mathrm{i}}\left(\mathrm{t}\mathrm{-}\mathrm{t\; \text{'}}\right)}{\mathrm{\delta }\left(\mathrm{t}\mathrm{-}\mathrm{t\; \text{'}}\right)}\mathrm{dt\text{'}}$$

with i = 1 and 2. The calculation of the temperature differences dT1 and dT2 thus differs only in the different time constants τ ₁ and τ ₂ .

und somit für das Faltungsintegral: $${\mathrm{dT}}_{\mathrm{i}}\left(\mathrm{t}\right)\mathrm{=}\frac{\mathrm{1}}{{\mathrm{\tau }}_{\mathrm{i}}}\underset{\mathrm{0}}{\overset{\mathrm{t}}{\mathrm{\int }}}\mathrm{mT}\left(\mathrm{tʹ}\right){\mathrm{e}}^{\mathrm{-}\frac{\mathrm{t}\mathrm{-}\mathrm{tʹ}}{{\mathrm{\tau }}_{\mathrm{i}}}}\mathrm{dtʹ}$$

mit 1=1 und 2.In the case of a simple thermodynamic model, as in the example FIG. 4 , the response function S (t) is given as a step response with the characteristic time constants τ ₁ and τ ₂ : $$\mathrm{S}\left(\mathrm{t}\right)\mathrm{=}\mathrm{1}\mathrm{-}{\mathrm{e}}^{\mathrm{-}\mathrm{t}\mathrm{/}{\mathrm{\tau }}_{\mathrm{i}}}$$

and thus for the folding integral: $${\mathrm{dT}}_{\mathrm{i}}\left(\mathrm{t}\right)\mathrm{=}\frac{\mathrm{1}}{{\mathrm{\tau }}_{\mathrm{i}}}\underset{\mathrm{0}}{\overset{\mathrm{t}}{\mathrm{\int }}}\mathrm{mT}\left(\mathrm{t\; \text{'}}\right){\mathrm{e}}^{\mathrm{-}\frac{\mathrm{t}\mathrm{-}\mathrm{t\; \text{'}}}{{\mathrm{\tau }}_{\mathrm{i}}}}\mathrm{dt\text{'}}$$

with 1 = 1 and 2.

At a constant sampling rate, the load drift compensation can also be used in this case, as in the calculation of the convolution integral in the preceding exemplary embodiment FIG. 4 , that convolutional integrals are calculated very efficiently by approximating them by a sum and calculating them in a recursive way.

Thus, the correction parameter CP is available in the form of the two temperature differences dT1 and dT2 for further calculation of a further, third correction sum C_DT.

The two temperature differences dT1 and dT2 can also be combined to form a single difference signal ΔdT according to the following formula: $$\mathrm{\Delta dT}=\mathrm{dT}2-\mathrm{dT}1=\left(\mathrm{ts}2-\mathrm{Te}\right)-\left(\mathrm{ts}1-\mathrm{Te}\right)=\mathrm{ts}2-\mathrm{ts}1.$$

The difference signal ΔdT thus represents the difference between two system temperatures, namely the difference between the second system temperature Ts2 and the first system temperature Ts1. Consequently, the correction parameter CP uses the difference signal ΔdT to characterize the difference between the two temperature differences dT1 and dT2. Finally, the correction parameter CP is then forwarded in the form of this difference signal ΔdT to the subsequent calculation module (shown in dashed lines).

FIG. 5c shows an exemplary time profile of the temperature differences dT1 and dT2 after a temperature jump dTe (shown in solid lines) acts on the balance. The temperature measurement signal mT, which largely corresponds to the temperature jump dTe, is shown in a simplified manner as a stepped solid line. While the temperature difference dT2 (shown in dotted lines) can follow the temperature jump dTe more quickly, the temperature difference dT1 (dashed line) reacts more slowly. The difference signal ΔdT, ie the difference between the two temperature differences: dT2-dT1, (dash-dotted line drawn) shows the observed behavior, namely an initial strong temperature influence, which subsides after some time.

als gute Approximation, sodass sich der dritte Korrektursummand C_DT gemäss folgender Formel bestimmt: $$\mathrm{C\_DT}\mathrm{=}\mathrm{c}\mathrm{*}\left(\mathrm{dT}\mathrm{1}\mathrm{-}\mathrm{dT}\mathrm{2}\right)\mathrm{*}\left(\mathrm{mL}\mathrm{+}\mathrm{m}\mathrm{0}\right)\mathrm{=}\mathrm{c}\mathrm{*}\mathrm{\Delta dT}\mathrm{*}\left(\mathrm{mL}\mathrm{+}\mathrm{m}\mathrm{0}\right)$$

wobei c und m0 Konstanten sind, die in geeigneter Weise zu bestimmen sind, beispielsweise durch Justierung.For the correction of the magnetic force f _M according to the above derivation applies: $$\mathrm{fm}\left(\mathrm{t}\right)\mathrm{-}\mathrm{fm}\left(\mathrm{0}\right)\mathrm{=}{\mathrm{a}}_{\mathrm{1}}\mathrm{\alpha }\frac{\left(\mathrm{dT}\mathrm{1}\mathrm{-}\mathrm{dT}\mathrm{2}\right)\left(\mathrm{L}\mathrm{+}\mathrm{m}\mathrm{1}\right)}{{\mathrm{b}}_{\mathrm{2}}\left(\mathrm{1}\mathrm{+}\mathrm{\alpha dT}\mathrm{2}\right)}\mathrm{\approx }\frac{{\mathrm{a}}_{\mathrm{1}}}{{\mathrm{b}}_{\mathrm{2}}}\mathrm{\alpha \; \Delta dT}\left(\mathrm{L}\mathrm{+}{\mathrm{<figure-callout\; id="1"\; label="m"\; filenames="imgf0001.png"\; state="\{\{state\}\}">m</figure-callout>}}_{\mathrm{1}}\right)$$

as a good approximation, so that the third correction sum C_DT is determined according to the following formula: $$\mathrm{C\_DT}\mathrm{=}\mathrm{c}\mathrm{*}\left(\mathrm{dT}\mathrm{1}\mathrm{-}\mathrm{dT}\mathrm{2}\right)\mathrm{*}\left(\mathrm{mL}\mathrm{+}\mathrm{m}\mathrm{0}\right)\mathrm{=}\mathrm{c}\mathrm{*}\mathrm{\Delta dT}\mathrm{*}\left(\mathrm{mL}\mathrm{+}\mathrm{m}\mathrm{0}\right)$$

where c and m0 are constants to be determined suitably, for example by adjustment.

Finally, this third correction sum C_DT is applied to the sum operator as in the case of the first correction or the second correction, which performs a corresponding correction by an additional addition of the third correction sum C_DT. The force measurement signal mL is thus additionally corrected with respect to dynamic changes in the ambient temperature Te. Again, the corrected for temperature differences output signal mLc is available for further processing.

FIG. 6 shows a block diagram in which the calculations according to FIG. 4 and 5 in that the calculation of the load drift correction and the calculation of the dynamic temperature behavior are combined as individual, essentially separate calculation modules. These modules are largely independent, and can be integrated into the overall calculation as needed. In this case, further modules can be interconnected in any number.

All of the correction weights, in this case the first correction sum C_S, the second correction sum C_LD for load drift correction, and the third correction sum C_DT for the correction of the dynamic temperature behavior are supplied to the sum operator. The addition of the three correction bands C_S + C_LD + C_DT then corrects the force measurement signal mL with respect to the influences described above. The corrected output signal mLc is forwarded for further processing, for example for forwarding to a display unit or to a further processing unit.

Furthermore, any combinations of the different measurement signals and temperature effects can also be calculated. For example, the short term increased temperature change of a component lead to different system temperatures in other components. For example, an abruptly increased magnet temperature, such as occurs when a load is applied to the force absorption, can be used to calculate the temperature changes dT1 and dT2 of the respective lever arms 48a and 48b and the associated change in length. Thus, the time-dependent temperature behavior can be calculated not only due to a change in the ambient temperature Te but also on the basis of the power loss P (t) in the coil, ie on the basis of the force measurement signal mL.

The inventive method for temperature correction and the inventive force measuring device 1 have been described and illustrated in preferred embodiments. The force measuring device has been described in the embodiment of a balance 1. However, the invention is also applicable to other force measuring devices, such as gravimetric measuring devices, weigh modules, load cells and measuring sensors, which may optionally form part of a balance. Of course, the devices are of course not limited to a particular choice, configuration, grouping and application of the components.

LIST OF REFERENCE NUMBERS

1

Force measuring device / balance

10

Load cell

11, 11a, 11b

Analog to digital converter

13

display unit

14

further processing unit

15

temperature sensor

20

Signal processing unit

41

boom

42

fixed area

43

vertically deflectable area

44

handlebars

45.47

flexure pivot

46

lever

48, 48a, 48b

lever arm

49

coupling element

50

permanent magnet

51

air gap

53

Kitchen sink

54

locator

56

control device

a1, a2

distance

C_S, C_LD, C_DT

correction summand

CP

correction parameters

dT1, dT2

temperature difference

dTe

temperature jump

dTr

relative temperature difference

F, F1, F2

transfer function

f _M

magnetic force

ic

compensating current

L

acting force / load

m, m1, m2

Dimensions

mL, mL '

Load signal

mT, mT '

Temperature measurement signal

mLc

output signal

P (t)

power loss

S (t), S1 (t), S2 (t)

response function

Ts, Ts1, Ts2

system temperature

Te

ambient temperature

tm

measured temperature

TK, TKLD

temperature coefficient

X

pivot point

α

expansion coefficient

β

relationship

ΔdT

difference signal

τ, τ ₁ , τ ₂

time constant

Claims (15)

Method for temperature correction of a force measuring device (1), in particular a balance, comprising the steps: - Generation of an electrical, corresponding to the applied force force measuring signal (mL) by means of a load cell (10); - measuring a temperature (Tm) which mainly corresponds to an ambient temperature (Te) acting on the force measuring device (1) and generating an electrical temperature measuring signal (mT) corresponding to the measured temperature (Tm); - processing of the force measurement signal (mL) based on the temperature measurement signal (mT) and the force measurement signal (mL) to a temperature-corrected output signal (mLc); Transmission of the output signal (mLc) to a display unit (13) and / or to a further processing unit (14),
characterized in that in the step of processing from the force measuring signal (mL) and / or the temperature measuring signal (mT), at least one of the correction of the output signal (mLc) serving correction parameter (CP) is calculated, which a temperature difference (dTr, dT1, dT2 ) characterized between: A) a system temperature (Ts, Ts1, Ts2) and the measured temperature (Tm) and / or B) a first system temperature (Ts1) and a second system temperature (Ts2)
consists. Method according to Claim 1, characterized in that the correction parameter (CP) is calculated by means of a response function (S (t), S1 (t), S2 (t)) and a convolution integral, in particular by the temperature measurement signal (mT), the force measurement signal ( mL) or a power loss (P) determined from the force measuring signal (mL) is convoluted with the time derivative of the response function (S (t), S1 (t), S2 (t)). A method according to claim 2, characterized in that the response function (S (t), S1 (t), S2 (t)) according to at least one thermodynamic law and / or as a thermodynamic response of an incoming step function or impulse function and / or as a time-decreasing Function, in particular an exponential function with a predetermined time constant (τ, τ ₁ , τ ₂ ) is defined. Method according to one of claims 1 to 3, characterized in that the correction parameter (CP) is calculated by a recursive calculation method. Method according to one of claims 1 to 4, characterized in that the correction parameter (CP) is calculated as a time-dependent variable, which depends on the course, in particular before the time change, the measured temperature (Tm) and / or the force measuring signal (mL). Method according to one of claims 1 to 5, characterized in that the measured temperature (Tm) by the system temperature (Ts, Ts1, Ts2) of a component of the force measuring device, in particular the load cell, is determined, said system temperature (Ts, Ts1, Ts2 ) corresponds mainly to the acting ambient temperature (Te). Method according to one of claims 1 to 6, characterized in that the system temperature (Ts, Ts1, Ts2) a temperature of a component of the force measuring device (1), in particular a power supply, a power transmission, a lever (46) or a lever arm (48). , and that the output signal (mLc) is corrected as a function of the temperature expansion of this component. Method according to one of claims 1 to 7, characterized in that the correction parameter (CP) at least two of the temperature differences (dT1, dT2) characterized, each of which different components of the force measuring device (1), in particular two counteracting lever arms (48a, 48b) assigned are. Method according to one of claims 1 to 8, characterized in that the correction parameter (CP) is calculated by means of a difference signal (ΔdT), which results from the difference between two temperature differences (dT1, dT2). Method according to one of claims 1 to 9, characterized in that the force measuring device (1) based on the principle of electromagnetic force compensation, and that the correction parameter (CP) a temperature difference is characterized, which is related to the generation of the compensation force. A method according to claim 10, characterized in that the compensation force by means of a current-carrying coil (53) to release a power loss (P) is generated, and that the correction parameter (CP) in dependence on the force measurement signal (mL), in particular a power loss derived therefrom ( P) is calculated. A method according to claim 10 or 11, characterized in that the compensation force is generated by the interaction of a coil (53) and a permanent magnet (50), and that the system temperature (Ts) a temperature of the coil (53) and / or the permanent magnet ( 50). Method according to one of claims 1 to 12, characterized in that at least two corrections of the force measurement signal (mL) by at least two different, substantially separate modules are calculated, each module by at least one correction parameter (CP) is calculated. Force measuring device (1), in particular a balance, with a load cell (10) which generates an electrical, corresponding to the applied force force measurement signal (mL), with a temperature sensor (15) which measures a temperature (Tm), the main one to the force measuring device acting ambient temperature (Te) and corresponds to the measured temperature (Tm) corresponding electrical Temperature measurement signal (mT) generated, and with a signal processing unit (20) which is configured to the force measurement signal (mL) based on the temperature measurement signal (mT) and the force measurement signal (mL) to a temperature corrected, to a display unit (13) and / or to process a further processing unit (14) transferable output signal (mLc), characterized in that the signal processing unit (20) is configured such that a correction parameter (CP) serving to correct the output signal (mLc) can be calculated, which is a temperature difference (dTr, dT1, dT2) between: A) a system temperature (Ts, Ts1, Ts2) and the measured temperature (Tm) and / or B) a first system temperature (Ts1) and a second system temperature (Ts2)
consists. Force measuring device (1) according to claim 14, characterized in that the temperature sensor (15) is in thermal contact with a, in particular fixed, component (40, 41, 42, 43) of the force measuring device (1), the system temperature (Ts, Ts1 , Ts2) of this component (40, 41, 42, 43) is determined mainly by the ambient temperature (Te).

Images